Non-intrusive Load Monitoring Based on Deep Convolutional Generative Adversarial Network Prediction

Abstract

Basic Non-Intrusive Load Monitoring is the field that encompasses energy disaggregation and appliance detection. In recent years, Deep Neural Networks have improved the classification performance, using the standard data representation that most datasets provide; that being low-frequency or high-frequency data. In this paper, we explore the NILM problem from the scope of generative adversarial network (GAN). We propose a way of changing the feature space with the use of an image representation of the data from EMBED dataset and the deep convolutional GAN (DCGAN). We then train some basic classifiers and use the acc score to test the performance of this representation. Multiple tests are performed to test the adaptability of the models to dierent appliances. We find that the performance average exceeds 70% in accuracy and in some cases that have more training epochs outperforms 80%.